Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Subject: Re: [PATCH v3 3/3] mtd: spinand: add SPI-NAND MTD resume handler
To:     Pratyush Yadav <p.yadav@ti.com>
CC:     Mark Brown <broonie@kernel.org>,
        Miquel Raynal <miquel.raynal@bootlin.com>,
        Vignesh Raghavendra <vigneshr@ti.com>,
        Boris Brezillon <boris.brezillon@collabora.com>,
        <linux-mtd@lists.infradead.org>,
        Alexandre Torgue <alexandre.torgue@foss.st.com>,
        <linux-spi@vger.kernel.org>,
        <linux-stm32@st-md-mailman.stormreply.com>,
        <linux-arm-kernel@lists.infradead.org>,
        <linux-kernel@vger.kernel.org>, <christophe.kerello@foss.st.com>
References: <20210527161252.16620-1-patrice.chotard@foss.st.com>
 <20210527161252.16620-4-patrice.chotard@foss.st.com>
 <20210528120508.f6viglv3gkzgweqq@ti.com>
 <3238725c-d7ff-c000-23d1-f18298e1556f@foss.st.com>
 <20210528131658.jtveyvdtxrwj2lgx@ti.com>
From:   Patrice CHOTARD <patrice.chotard@foss.st.com>
Message-ID: <e4722cab-ab76-137b-9224-43d4c68d013a@foss.st.com>
Date:   Fri, 28 May 2021 16:31:59 +0200
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Thunderbird/68.10.0
MIME-Version: 1.0
In-Reply-To: <20210528131658.jtveyvdtxrwj2lgx@ti.com>
Content-Type: text/plain; charset="utf-8"
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Precedence: bulk


On 5/28/21 3:17 PM, Pratyush Yadav wrote:
> On 28/05/21 02:51PM, Patrice CHOTARD wrote:
>> Hi Pratyush
>>
>> On 5/28/21 2:05 PM, Pratyush Yadav wrote:
>>> On 27/05/21 06:12PM, patrice.chotard@foss.st.com wrote:
>>>> From: Patrice Chotard <patrice.chotard@foss.st.com>
>>>>
>>>> After power up, all SPI NAND's blocks are locked. Only read operations
>>>> are allowed, write and erase operations are forbidden.
>>>> The SPI NAND framework unlocks all the blocks during its initialization.
>>>>
>>>> During a standby low power, the memory is powered down, losing its
>>>> configuration.
>>>> During the resume, the QSPI driver state is restored but the SPI NAND
>>>> framework does not reconfigured the memory.
>>>>
>>>> This patch adds SPI-NAND MTD PM handlers for resume ops.
>>>> SPI NAND resume op re-initializes SPI NAND flash to its probed state.
>>>>
>>>> Signed-off-by: Christophe Kerello <christophe.kerello@foss.st.com>
>>>> Signed-off-by: Patrice Chotard <patrice.chotard@foss.st.com>
>>>> ---
>>>> Changes in v3:
>>>>   - Add spinand_read_cfg() call to repopulate cache
>>>>
>>>> Changes in v2:
>>>>   - Add helper spinand_block_unlock().
>>>>   - Add spinand_ecc_enable() call.
>>>>   - Remove some dev_err().
>>>>   - Fix commit's title and message.
>>>>
>>>>  drivers/mtd/nand/spi/core.c | 33 +++++++++++++++++++++++++++++++++
>>>>  1 file changed, 33 insertions(+)
>>>>
>>>> diff --git a/drivers/mtd/nand/spi/core.c b/drivers/mtd/nand/spi/core.c
>>>> index 1f699ad84f1b..e3fcbcf381c3 100644
>>>> --- a/drivers/mtd/nand/spi/core.c
>>>> +++ b/drivers/mtd/nand/spi/core.c
>>>> @@ -1099,6 +1099,38 @@ static int spinand_block_unlock(struct spinand_device *spinand)
>>>>  	return ret;
>>>>  }
>>>>  
>>>> +static void spinand_mtd_resume(struct mtd_info *mtd)
>>>> +{
>>>> +	struct spinand_device *spinand = mtd_to_spinand(mtd);
>>>> +	int ret;
>>>> +
>>>> +	ret = spinand_reset_op(spinand);
>>>> +	if (ret)
>>>> +		return;
>>>> +
>>>> +	ret = spinand_read_cfg(spinand);
>>>> +	if (ret)
>>>> +		return;
>>>> +
>>>> +	ret = spinand_init_quad_enable(spinand);
>>>> +	if (ret)
>>>> +		return;
>>>> +
>>>> +	ret = spinand_upd_cfg(spinand, CFG_OTP_ENABLE, 0);
>>>> +	if (ret)
>>>> +		return;
>>>> +
>>>> +	ret = spinand_manufacturer_init(spinand);
>>>> +	if (ret)
>>>> +		return;
>>>> +
>>>> +	ret = spinand_block_unlock(spinand);
>>>> +	if (ret)
>>>> +		return;
>>>> +
>>>> +	spinand_ecc_enable(spinand, false);
>>>> +}
>>>> +
>>>
>>> I don't think you quite get what me and Miquel are suggesting.
>>>
>>> The helper should call all these functions like read_cfg() 
>>> quad_enable(), etc. So it should look something like:
>>
>> Yes, this series was sent too quickly on my side, and i misunderstood 
>> what you suggested, sorry for that.
> 
> Ah, I thought this was the v4. I see Miquel clarified already in reply 
> to this patch.
> 
>>
>>>
>>> int spinand_init_flash()
>>> {
>>> 	ret = spinand_read_cfg(spinand);
>>> 	if (ret)
>>> 		return;
>>>
>>
>> The new helper spinand_read_cfg() must not be called in spinand_init_flash()
>> but directly in spinand_resume().
>> This because spinand_read_cfg() only performs a REG_CFG read without doing the
>> memory allocation of spinand->cfg_cache.
>>
>> In spinand_init(), spinand_init_cfg_cache() must be called as previously as it does spinand->cfg_cache
>> memory allocation and call the new helper spinand_read_cfg(). 
>> Then after, spinand_init_flash() can be called.
> 
> I think it would be simpler for spinand_init_cfg_cache() to just 
> allocate the spinand->cfg_cache array and then for spinand_init_flash() 
> to call spinand_read_cfg().

Argh, i just send the v5 ;-) .... i will wait the full v5 review and will follow your proposal 
regarding spinand_init_cfg_cache().

Patrice

> 
>>
>>> 	ret = spinand_init_quad_enable(spinand);
>>> 	if (ret)
>>> 		return;
>>>
>>> 	ret = spinand_upd_cfg(spinand, CFG_OTP_ENABLE, 0);
>>> 	if (ret)
>>> 		return;
>>>
>>> 	ret = spinand_manufacturer_init(spinand);
>>> 	if (ret)
>>> 		return;
>>>
>>> 	ret = spinand_block_unlock(spinand);
>>> 	if (ret)
>>> 		return;
>>>
>>> 	spinand_ecc_enable(spinand, false);
>>> }
>>>
>>> Then spinand_mtd_resume should look something like:
>>>
>>> int spinand_mtd_resume()
>>> {
>>> 	ret = spinand_reset_op(spinand);
>>> 	if (ret)
>>> 		return;
>>>
>>> 	return spinand_init_flash();
>>> }
>>>
>>> And spinand_init() should look something like:
>>>
>>> int spinand_init()
>>> {
>>> 	...
>>> 	spinand->oobbuf = ...
>>>
>>> 	spinand_init_flash();
>>>
>>> 	spinand_create_dirmaps();
>>>
>>> 	...
>>
>> As explained just above, spinand_init() will look like :
>>
>> int spinand_init()
>> {
>> 	...
>> 	spinand->oobbuf = ...
>>
>> 	spinand_init_cfg_cache();     => perform cfg cache memory allocation and read the REG_CFG
>>
>> 	spinand_init_flash();
>>
>> 	spinand_create_dirmaps();
>>
>>> }
>>>
>>>
>>>>  static int spinand_init(struct spinand_device *spinand)
>>>>  {
>>>>  	struct device *dev = &spinand->spimem->spi->dev;
>>>> @@ -1186,6 +1218,7 @@ static int spinand_init(struct spinand_device *spinand)
>>>>  	mtd->_block_isreserved = spinand_mtd_block_isreserved;
>>>>  	mtd->_erase = spinand_mtd_erase;
>>>>  	mtd->_max_bad_blocks = nanddev_mtd_max_bad_blocks;
>>>> +	mtd->_resume = spinand_mtd_resume;
>>>>  
>>>>  	if (nand->ecc.engine) {
>>>>  		ret = mtd_ooblayout_count_freebytes(mtd);
>>>> -- 
>>>> 2.17.1
>>>
>>
>> Thanks
>> Patrice
>